FROZEN SHOULDER - A LONG-TERM FOLLOW-UP

Sixty-two patients (sixty-eight shoulders) who had been treated non-operatively for idiopathic frozen shoulder were evaluated subjectively and objectively at two years and two months to eleven years and nine months of follow-up (average, seven years). Thirty-one (50 per cent) of these patients still had either mild pain or stiffness of the shoulder, or both. The range of motion averaged 161 degrees of forward flexion, 157 degrees of forward elevation, 149 degrees of abduction, 65 degrees of external rotation, and internal rotation to the level of the fifth thoracic spinous process. Thirty-seven (60 per cent) of the sixty-two patients still demonstrated some restriction of motion as compared with study-generated control values (calculated as the average motion, in each plane, for the thirty-seven unaffected shoulders of the patients who had unilateral disease). Ten patients had restriction of forward flexion; eight, of forward elevation; seventeen, of abduction; twenty-nine, of external rotation; and ten, of internal rotation. However, when the motion of each affected shoulder of thirty-seven patients who had unilateral involvement was compared with that of the unaffected contralateral shoulder, eleven (30 per cent) demonstrated some restriction. None of these patients had restriction of forward flexion; two had restriction of forward elevation; two, of abduction; seven, of external rotation; and seven, of internal rotation. The patients who had substantial restriction in three planes or more were thirteen times more likely to be men (p > 0.05). Marked restriction, when it was present, was most commonly in external rotation. Only seven patients (11 per cent) reported mild functional limitation. There was no statistical correlation between the objective outcome (the range of motion and radiographic findings) and the subjective findings (pain, stiffness, and functional restriction). There was, however, a strong statistical association between functional limitation and the presence of symptoms (p < 0.05). Neither the subjective nor the objective outcome was statistically related to age, whether the dominant or the non-dominant extremity was affected, whether the left or right shoulder was affected, the nature of the onset of symptoms, the etiology, the method of treatment, bilateral involvement, or associated medical conditions. Although most patients had no symptoms or functional restriction, at the time of the long-term follow-up thirty-seven (60 per cent) of the sixty-two patients had measurable restriction when compared with control values, and eleven (30 per cent) of thirty-seven unilaterally affected patients had measurable restriction when the involved shoulder was compared with the contralateral, unaffected shoulder.